AF 447 Search to resume

[PJ2]

Lonewolf;

I'll guess that the fuel tanks suffered trauma

I think it is reasonable to posit that the wing surfaces would have separated as a result of a "hydrodynamic pressure surge" from the [incompressible] fuel, (discussed in the BEA Report on the Concorde accident, Appendix 9, in French here, and on pg. 111 of English version of the entire report here), and relatively little air-space in the outer and inner tanks, and the horizontal stabilizers would likely behave the same way as they would have been full.

grity...very imaginative work - I was also examining it using various color filters to see what showed in terms of patterns, etc...

One thing about stretching...we have the vertical and horizontal scale so we have a ratio...does stretching the image maintain the ratio, I wonder? It's essentially 1:3...

[mm43]

Originally posted by grity ...
the main difference between falling in air and in water is the specific weight of the medium, and it goes into the speed formula with his square root.....
water is 1000 times heavier than air, sqrt 1000= 31.6

so if it was fallen through the air with 75 m/s it will sink under water with a dimension of 75/31,6=2,4 m/s and so it will need around 4000/2,4=1666 sec (half an hour!) to reach the sea bottom

Technically true, but Archimedes' principle gets involved at the air/water interface. As I have pointed out in previous posts there are reciprocal buoyancy moments that interact at the time of impact, and even during the course of a break-up, there will be items that are compressible and others that aren't. For instance, the moment the originally intact hull is breached, the total buoyancy moment is lost, but within the hull there are numerous items that retain buoyancy - not to forget the fuel.

This aircraft has broken up at the surface after a high AoA and high rate of descent impact. Some items, e.g. engines will part company with the pylon mounts immediately due to the obvious disparity in their mass and volume when compared to the intact area of the wing to which they were attached. The engines actual impact velocity will be further reduced by the difference between the acceleration force that stops the wing and the force that shears their mountings.

All that means is that it is not a simple matter to determine at what rate an object will sink, as it will depend on many factors that make up a homogenous mass. Some items with profiles that present "slippery" surfaces in a fluid dynamics sense may take trajectories that their shape permits, while other stuff will just "sink".

As has previously been said, there will be no rush to the bottom, as increasing pressure equals increasing surface friction.

There will be a relationship to where items are located on the bottom and the time that it took them to get there. A small subsurface current heading west will be responsible for some of the east/west distribution.

In the case of the HMS "Ark Royal", it would appear that it broke in two on the surface, and the two separate halves then had differently distributed centers of gravity and fluid dynamics allowed them to head to the bottom on differing trajectories. Remember that a ship that no longer floats will still have intact water-tight compartments that will impact on its center of buoyancy and cause its attitude to change during the sinking process. Hence a possible reason the two halves were found so far apart.

[HazelNuts39]

Lonewolf;

I'll guess that the fuel tanks suffered trauma

Look also at what appears to be wing lower skin attached to MLG, with tank access holes.

[HazelNuts39]

Originally Posted by grity

the main difference between falling in air and in water is the specific weight of the medium

Just to keep it simple, in water as in air, three forces and their moments are at play on a free-falling body: Weight, buoyancy and fluid dynamic force. Buoyancy is equal to the weight of the displaced volume of the medium, usually negligible in air but not in water. The fluid dynamic force is equal to fluid density times velocity-squared times a factor that depends on the geometry of the moving body: dimensions, shape and orientation relative to the direction of movement.

[auv-ee]

Originally Posted by mm43

As has previously been said, there will be no rush to the bottom, as increasing pressure equals increasing surface friction.

Generally agreed with your post, except this statement. Perhaps you meant something else. The density of water changes very little (some, but not much) with increasing pressure. The form-drag (in the applicable regime) is proportional to the density of the media, and no other property of it. Thus drag is not a significant function of depth.

Edit: mm43 did mention skin drag in his post, and I believe that involves the viscosity of the media. I found a reference (Water Viscosity at High Pressure and Temperature) that shows the viscosity to double at 1Gpa, roughly equivalent to 4000m depth. However, I think the drag of these A/C parts is dominated by form drag. I'm definitely getting out of my area here, but I've never heard the MEs I've worked with factor depth into drag calculations.

--

I might point out that objects that tend to "kite" off the vertical during descent will also be likely to turn. Thus most objects will still end up near each other after either a straight or spiral descent. The example of the Ark Royal differs in that symmetry and mass will make the turn rate very slow, and thus a hull section could wind up a long way away. It is the rare object that will drift away on a constant heading.

[Shadoko]

Hi,

About the meanings of the wreckage sonar picture : "The most notable discovery the team made was that the ship had split apart, the stern section lying 1,970 feet (600 m) from the bow section and facing opposite directions." (RMS Titanic - Wikipedia, the free encyclopedia). Titanic. A ship, standing (not enough...) at the surface. No speed. No big "splash". Sea like a mirror, as it is known. About the same depth. So ...

[HazelNuts39]

Originally Posted by grity

what can be the struktur right forward of the wing? part of the wing or part of the fuselage?

Just a guess: I think we are looking at the upper skin of the left wing tip, with aileron hinge brackets. The structure in the background could be a slat.

[mm43]

Originally posted by auv-ee ...
Perhaps you meant something else.

Perhaps I did!

An object sinking under the influence of gravity is subject to a resistance/drag force which increases with velocity. A maximum velocity is reached when the drag force equals the propelling force. This constant velocity is the "terminal velocity". There is also a case where objects are moving at low speeds and where turbulence is a minor factor, that viscous drag determines the "terminal velocity".

So in a sense, you can "have your cake and eat it to". In the original rather crass statement, I should really have said that objects will reach a terminal velocity rather implying that pressure and viscous drag were directly related.

[auv-ee]

Originally Posted by Shadoko

...the stern section lying 1,970 feet (600 m) from the bow section and facing opposite directions." (RMS Titanic - Wikipedia, the free encyclopedia). Titanic. A ship, standing (not enough...) at the surface. No speed. No big "splash". Sea like a mirror, as it is known. About the same depth. So ...

This phenomenon was just discussed in posts 3550 and 3553, above, and is, I believe, applicable to the Titanic.

[Machinbird]

AUV-EE comment

I might point out that objects that tend to "kite" off the vertical during descent will also be likely to turn. Thus most objects will still end up near each other after either a straight or spiral descent.

Good comment! The MLG strut probably came down like a helicopter after the surface breakup due to the attached skin panel. Any angle of attack of that 'vane' would cause a rotation. Probably accounts for the relative angle of the panel and gear (appearance of being extended).

Are we agreed that the subsurface current dispersing the wreckage was in a East to West direction?

The dispersion of this wreckage reminds me of an accident where the fuel washed components of the fractured forward fuselage out of the crater and they were dispersed with port side components further from the crater than starboard side components but spread out along a narrow angle corresponding to the distance off axis in the aircraft. In that case, it was easy to find components by observing what was in the vicinity.

With AF447 components being sorted by starting position, and fall rate, it is a much more difficult problem to figure out where to find things. But if the recorders are as dense as others have indicated and if they have torn loose, at least they should be close to other items of similar density (as modified by their starting position).

I expect that the recovery team will go for the tail section first to attempt to locate the recorders there. If they are no longer there however, they had best be careful because once they disturb the 'indian signs' on the bottom, they will have to locate the recorders by means other than visual.

AF447 is a bit more broken up than I was hoping for. I expect one of the reasons that the pingers were not located is that they are buried and the bottom muck attenuated the signal.
I expect people at BEA and the recovery team are going over the wreckage pictures trying to develop an initial orientation for the aircraft so they can make some educated guesses where things can be found.

[auv-ee]

Originally Posted by Machinbird

Are we agreed that the subsurface current dispersing the wreckage was in a East to West direction?

That is the way the image looks to me. I don' t know if that interpretation is supported by any other current data. At this point, the debris dispersion may be the only evidence of the actual current at that time and place. It may be that no one really knows the current by any other means, because the measured data are so sparse and the current is so variable.

[mm43]

Originally posted by Machinbird ...

Are we agreed that the subsurface current dispersing the wreckage was in a East to West direction?

In the Shom PDF document on page 5, there is a current chart showing that at 1,000m the current is west at about 8cm/sec. I suspect possibly a little slower at 4,000m. Photographs of debris show no sign of any sand scouring and it looks very still on the bottom.

[grity]

I think we are looking at the upper skin of the left wing tip, with aileron hinge brackets. The structure in the background could be a slat.

HazelNuts39, it is not only the tip, it is the middle part of the left wing "...NOT WALKE OUTSIDE THIS ARE...." in this area are spoiler not ailerons
but which slat has such a complex struktur? grity

[JD-EE]

Chris Scott, please leave the "stupid boy" off. I try to respect people more than that. (As one slightly your senior I can't even pat you on the head and utter words about "silly boy".)

(I might note you did get me to use my modest amount of gray mush to look again at the debris field size and figure out what it probably amounts to, which is not the plane's heading.)

[Shadoko]

Hi,

According to the relative sizes of the "DO NOT WALK OUTSIDE THIS AREA" writings and the wing width, it seems that quite half the front part of the wing is missing. Or bent downward? Is this compatible with a belly strike, nose up? In other way, all mobile surfaces seem ripped out. What attitudes at impact could be compatible with both? Very high speed, anyway, no?

[Machinbird]

Outliers

There are several elements that are outliers in the scan image. By outliers, I mean wreckage lying at a greater distance from the mean wreckage centerline than can be explained by their starting location on the aircraft. I am reposting MM43's image below:



Possible explanations for these outliers are:

• Thrown away from the aircraft during the initial crash and flew through the air for some distance.
• Floated on the surface for some time before sinking.
• Glided in a very slow turning stable configuration on its way to the bottom.

Identification of these outliers and their configuration should answer which influence applied. The elements that could have been thrown have particular interest in that part of the aircraft's velocity vector can be imputed from their location (but not directly from its final position). It is still necessary to estimate its time of flight in the water column.

Probably the best way to impute the aircraft's final heading is by drawing a line normal to the locations of the two engines. It won't be perfectly accurate, but probably within 20 degrees of actual due to dispersion probabilities. The heading/reciprocal heading should be inferred by relative position of various wreckage elements or by actually ID of a particular engine.

[CONF iture]

Originally Posted by sensor_validation

Transportation Safety Board of Canada - AVIATION REPORTS - 2008 - A08W0007

Very informative report by the canadian TSB.
A good reminder how inputs on flight controls at cruising flight levels must be moderate.

Also of interest :

The automatic reversion to alternate law due to differences in computed angle of attack values was also a product of pilot control inputs.

[Mr Optimistic]

There are a number of 'objects' which are doubled up, ie have a parallel twin just to one side (including the northern most outlier). Presume these are in fact images one of another where the two scans didn't quite align when the composite was constructed ?

[HarryMann]

Originally Posted by auv-ee

Generally agreed with your post, except this statement. Perhaps you meant something else. The density of water changes very little (some, but not much) with increasing pressure. The form-drag (in the applicable regime) is proportional to the density of the media, and no other property of it. Thus drag is not a significant function of depth.

Well, you might think that, but I couldn't possibly comment...

But will, anyway.

If the viscosity changes significantly with depth (x 2), then Reynolds number does too (for a given free-fall speed).
Form (profile) drag definitely changes with Re, as Re determines the flow regime* (hint: dimples in golf ball make large difference, they simulate a much higher Re)

Thus, I'd say, even profile drag is depth dependent, although one would have to iterate to get the terminal speed (Re not invariant with speed), which may go up or down with >> Viscosity. Not going to predict that here..

Agree that profile drag (form drag) is an order of magnitude more than skin friction, but this would still change somewhat


* e.g. Laminar separation bubble; vortex wake; fully attached (inviscid); fully separated flow, etc

NB. This effect is either negligible, small or if Re near a critical flow regime change (Re crit) > then could make quite a big difference to drag.
Streamlined shapes (auvs en vol de ligne!) at a particular (low) speed range could well, as you say, have effectively depth independent drag.
But certainly not every object at any speed...

[Chris Scott]

A319 Wake-Turbulence Encounter

Salut Conf iture,

I think that sensor_validation and you raise a very relevant point about this serious incident or accident in 2008, which was exacerbated by the PF’s over-control in roll and yaw. The TSB of Canada’s report makes very interesting reading. I’m going to repeat the sentence you quoted:

“The automatic reversion to alternate law due to differences in computed angle of attack values was also a product of pilot control inputs.”

Here are some passages from the report’s description of the way the aircraft was handled after the PF disconnected the AP in unexpected wake turbulence, which he apparently mistook for AP misfunction.(Any highlighting is mine.)

Quote:
"During the 18-second duration of the event, heading varied from 065ºM to 086ºM. The captain reacted to the rolls with a total of nine sidestick roll inputs, accompanied by coordinated rudder pedal deflections. Five sidestick inputs were to full travel of 20º."

Quote:
"From 06:48:07 to 06:48:25, pilot sidestick roll inputs were 90º out of phase with aircraft motion. From 06:48:07 to 06:48:15, lateral accelerations and heading deviations were approximately 90º out of phase with the rudder pedals. This indicated that after the autopilot was disconnected, most of the aircraft motion in the roll axis resulted from pilot inputs and that lateral accelerations were due mostly to pilot rudder control inputs."

Quote:
“Abnormal accelerations in the normal or vertical axis were correlated with changes in angle of attack, and sidestick pitch control inputs opposed these aircraft angle of attack excursions.”

I think the expression “sidestick pitch control inputs” may have been unintended, because it goes on to say:
“...the accelerations in the vertical were a result of external influences on angle of attack associated with wake turbulence,”
　
In March 2008 – long before the issue of the TSB of Canada’s report – PPRuNe (as CONF iture, PJ2 and others will remember) was discussing at length a rough landing by an A320 in a limiting crosswind at Hamburg. The subject of SIDESTICK technique came up when someone gave a link to this

video" target="_blank">video

of a pilot using the sidestick during a crosswind landing. (N.B. This was NOT a video of the Hamburg incident!) One poster described it as coffee stirring. Having seen it before from the jump seat during 12 years’ line-checking, I called it “sidestick abuse” in this post:
http://www.pprune.org/tech-log/31609...ml#post3979423

[SIZE=2][FONT=Verdana]As for the use of "co-ordinated" RUDDER at cruise speeds, I would argue that it was nothing of the kind: the report suggests that, notwithstanding the rudder-travel limiter, A320-family pedal loads need to be increased. (No comment re the A330.) Should rudder ever be used in cruise-flight, apart from the asymmetry case? I remember being disturbed by one copilot, who moved a pedal significantly when stretching a leg that had "gone to sleep".

In view of the possibility that AF447 encountered sudden, severe turbulence, perhaps we should be adding another scenario to the list of possible causes of LOC and Control-Law degradation. Could it be compatible with the analysis of ACARS messages? I think it might be.